Binary Imbalanced Data Classification Based on Modified D2GAN Oversampling and Classifier Fusion
Open Access
- 14 September 2020
- journal article
- research article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 8, 169456-169469
Abstract
Binary imbalance problem refers to such a classification scenario where one class contains a large number of samples while another class contains only a few samples. When traditional classifiers face with imbalanced datasets, they usually bias towards majority class resulting in poor classification performance. Oversampling is an effective method to address this problem, yet how to conduct diversity oversampling is a challenge. In this paper, we proposed a diversity oversampling method based on a modified D2GAN model, and on the basis of diversity oversampling, we also proposed a binary imbalanced data classification approach based on classifier fusion by fuzzy integral. Extensive experiments are conducted on 8 data sets to compare the proposed methods with 7 state-of-the-art methods on 5 aspects: MMD-score, Silhouette-score, F-measure, G-means, and AUC-area. The 7 methods include 4 SMOTE related approaches and 3 GAN related approaches. The experimental results demonstrate that the proposed methods are more effective and efficient than the compared approaches.Keywords
Funding Information
- Key Research and Development Program of Science and Technology Foundation of Hebei Province (19210310D)
- Natural Science Foundation of Hebei Province (F2017201026)
- Postgraduate’s Innovation Fund Project of Hebei University (hbu2019ss077)
This publication has 35 references indexed in Scilit:
- Classification of Imbalanced Data by Oversampling in Kernel Space of Support Vector MachinesIEEE Transactions on Neural Networks and Learning Systems, 2017
- The classification of imbalanced large data sets based on MapReduce and ensemble of ELM classifiersInternational Journal of Machine Learning and Cybernetics, 2015
- A Hierarchical Pattern Learning Framework for Forecasting Extreme Weather Events2015
- Improved SMOTEBagging and its application in imbalanced data classification2013
- DBSMOTE: Density-Based Synthetic Minority Over-sampling TEchniqueApplied Intelligence, 2011
- Learning from Imbalanced DataIEEE Transactions on Knowledge and Data Engineering, 2009
- Safe-Level-SMOTE: Safe-Level-Synthetic Minority Over-Sampling TEchnique for Handling the Class Imbalanced Problem2009
- Extreme learning machine: Theory and applicationsNeurocomputing, 2006
- The Nature of Statistical Learning Theory1995
- Silhouettes: A graphical aid to the interpretation and validation of cluster analysisJournal of Computational and Applied Mathematics, 1987